Multi-Pattern Resource Allocation in Heterogeneous Cellular Networks

Interference is a fundamental limiting factor in throughput performance in a heterogeneous cellular network due to resource sharing between tiers of base stations. To cope with this issue, novel resource allocation strategies are needed to jointly optimize user association, multicell multiuser channel assignment and power control. However, the interference coupling introduces the inherent non-convexity in the upcoming optimization problems, preventing researchers from obtaining globally optimal solution for a reasonably-sized network. In this paper, we propose a new framework for multicell resource allocation to consider the interference coupling. Specifically, we characterize the interference coupling by pre-defining interference patterns, and perform the resource allocation across these patterns. One key feature of this interference pattern formulation is that the patterns remain fixed and independent of the optimization process. This creates a favorable opportunity for convex formulations. By specifying all possible patterns, we derive the optimal benchmark of the system performance that can be achieved by network-wide resource optimization. Our framework also enables us to analyze and compare a wide range of resource allocation strategies that have been developed previously in very diverse contexts in a unified view, which is done by restricting the candidate interference patterns to a certain set of pre-defined ones. Finally, we provide the numerical results of performance evaluation of various strategies.